I’m currently working on task of learning representation (deep embeddings). The dataset I use have only one example image per object. I also use augmentation.
During training, each batch must contain N different augmented versions of single image in dataset (dataset[index] always returns new random transformation).
Is there some standard solution or library for this purpose, that will work with torch.utils.data.distributed.DistributedSampler?